THE BENEFITS OF TREE-RING CURVES DETRENDING FOR DATING ARCHAEOLOGICAL WOOD

During the process of developing the standard chronology for oak for the Czech Republic, two versions of standard chronology were determined: the standard chronology developed using detrended tree-ring series and the standard chronology developed using non-detrended tree-ring series. These standard chronologies were applied to date detrended and non-detrended average tree-ring curves obtained from dendrochronological samples from selected archaeological locations. The highest values of the t-test were achieved when comparing the detrended or non-detrended average ring curves only and exclusively with the detrended standard chronology. Similarly, the highest percentage of the curve parallelism was always obtained in comparison with the detrended standard chronology. The confidence and, more importantly, the prediction intervals of the detrended ring curves are considerably smaller than those of the non-detrended ones. The regression curves of the detrended standardised values of the ring curves are more similar to the calibration curve than the non-detrended ones. The significance level of the regression models in the detrended ring curves is notably higher, than in the case of the non-detrended ring curves. The differences established between the detrended and non-detrended standard chronology have highlighted the importance to develop detrended standard chronologies.